We propose a longitudinal overlapping cohort study of adolescent behavioral development, using juvenile twins and their parents. To provide a basis for understanding the causes of adolescent psychopathology, we will resolve genetic and environmental influences on behavioral development. We have assembled the population of around 7000 school age twin pairs known to the education system in Virginia. From this population we will recruit 100 juvenile twin pairs at each age 8 through 16. In each subsequent year a new cohort of 8 year olds will be recruited. The twins and their parents will be interviewed at home each year for four years; the twins, their parents and their teachers will provide supplementary questionnaire information. We will assess: conduct and delinquency, depression, attention deficit and hyperactivity; familial and extrafamilial environment risk factors; stage of puberty and drug use. The overlapping cohort feature of our design avoids some of the shortcomings of longitudinal research. The genetic information provided by the twin study aspects of the design, augmented with information from the parents, will enable us to test a series of cross sectional and developmental hypotheses using both univariate and multivariate analyses. The inclusion of opposite sex twin pairs will give important additional information about the development of sex differences during adolescence. For example, we will: test alternative hypotheses about the causes of the increase in depression, and the development of the sex difference in depression, during adolescence; examine the sources of the apparent increases in environmental effects may be relatively transient; test alternative explanations for the observed co-morbidity of conduct disorder and depression; examine alternative explanations for the changing pattern of familial resemblance for affective disorders during development.